面向智能交通的图像去雾技术的实现

Implementation of Image Fog Removal Technology for Intelligent Transportation

摘要: 雾，众所周知，其在很大程度上会阻碍我们看清楚生活中的现象，如果在这种情况下拍照的话，自然而然对比度肯定会不如正常情况下，这样就给我们在实际生活中带来很多不便。特别是在交通这一方面，由于此天气现象的存在，可能没有办法很清楚的看到行车实况，这种情况造成的一个比较小的影响就是车与车的摩擦，更严重的一点的话就是车祸了。那么，对这种自然现象造成的图像不清晰展开图像信号处理与研究是非常有必要的。因此，通过调研我们发现Matlab能够将在恶劣天气条件下获得的图片进行处理，使其更加接近原始图像。这对于上述问题是一个很好的解决工具。本次研究具体使用了三种算法。分别是局部直方图均衡化，全局直方图均衡化还有就是Retinxe算法。而且这项技术的出现，对于我们处理别的不清晰的图像问题也是非常有帮助的。

Abstract: Fog, as we all know, can prevent us from seeing things clearly in our daily life to a large extent. If we take photos under such circumstances, the contrast ratio will certainly be lower than under normal circumstances, which will bring us a lot of inconvenience in real life. Especially in the aspect of traffic, due to the existence of this weather phenomenon, there may be no way to clearly see the driving conditions. A relatively small impact caused by this situation is the friction between cars. A more serious point is the car accident. Therefore, it is necessary to carry out image signal processing and research on the image blurring caused by this natural phenomenon. Therefore, we found that Matlab was able to process images obtained under haze conditions to bring them closer to the original image. Three algorithms are used in this study. They are local histogram equalization, global histogram equalization, and Retinxe algorithm. And the advent of this technology is also very helpful for us to deal with the problem of other images that are not clear .